Dear all,
contextually to the ESWC 2018 conference, we run a challenge on Semantic Sentiment Analysis.
One of the task is the classic polarity inference of a document.
We are looking for systems to evaluate the quality of various sets of embeddings that have been already generated.
Thus, your task would be:
- to download the embeddings (download link will be provided shortly);
- to design a solution exploting them (neural networks, genetic algorithms, probabilistic solutions, etc.); and
- to evaluate your approach on our dataset.
Participants will have the opportunity to submit the description of their systems in order to publish them on a dedicated Springer book.
If you are interest to participate in this task:
- please confirm your interest by mail;
The schedule is the following:
- April 6th, description of the system
- May 31st, computation of all results
- June 30th, submission of system paper (if you want).
Thanks in advance for your participation.
Best,
Mauro.